Metraipolis › archive › dc_3a2880964482
Tell Robot What Not to Do: A Negation Understanding Perspective — arXiv:2610.11952
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.11952
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-09T06:35:22.763Z
- SHA-256 of the archived text
de467cea8fe390610ef51d9e0da21e9212dc922c02adedcb2246ff468b0cac68- Stable URL of this page
https://metraipolis.com/documento/dc_3a2880964482
How to cite
Tell Robot What Not to Do: A Negation Understanding Perspective — arXiv:2610.11952 — arXiv (en), https://arxiv.org/abs/2610.11952 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3a2880964482 (Metraipolis archive dc_3a2880964482, sha256 de467cea8fe390610ef51d9e0da21e9212dc922c02adedcb2246ff468b0cac68)
BibTeX
@misc{dc3a2880964482,
title = {Tell Robot What Not to Do: A Negation Understanding Perspective — arXiv:2610.11952},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.11952},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3a2880964482; sha256 de467cea8fe390610ef51d9e0da21e9212dc922c02adedcb2246ff468b0cac68},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11952. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Instruction following enables robots to perform diverse tasks specified in natural language, making it a fundamental capability for human-robot interaction. Beyond communicating desired outcomes, users also need to specify constraints on what not to do. We investigate how to enable vision-language-action models (VLAs) to follow negated instructions, where robots must accomplish task goals while respecting explicit exclusions. To this end, we propose NegaAlign, a parameter-efficient, plug-and-play framework that extends pretrained VLAs to follow negated instructions through image-language supervision alone. Specifically, we introduce Negation Transformation Layers into selected layers of the vision-language backbone to reshape intermediate instruction representations. Meanwhile, a teacher-guided alignment mechanism is designed to align instruction-relevant visual tokens, transferring action-relevant grounding from instructions that satisfy the negated constraint. The training phase uses supervision constructed from existing demonstrations and updates only the inserted layers, keeping all pretrained parameters frozen, including the action generator. We further introduce NegaBench, a simulation benchmark spanning 10 scenarios across five domains for systematically evaluating manipulation under negated constraints. Experiments across GR00T, $\pi_0$, and $\pi_{0.5}$ demonstrate consistent improvements in negated instruction following. With 11.6M trainable parameters, NegaAlign increases the negated-instruction success rate of $\pi_{0.5}$ from 2.60% to 88.45% on NegaBench and from 12.4% to 88.8% on real-world tasks, while retaining performance on affirmative instructions.
Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.RO, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO).
Pagina del lavoro: https://arxiv.org/abs/2610.11952 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-09T05:55:22.188Z dalla pagina https://arxiv.org/list/cs.RO/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).